There is a growing concern over the rising incidence of nonmelanoma skin cancer (NMSC), and higher treatment-related cost may become a burden on the healthcare system. Management of NMSC is usually surgical either by local excision or Mohs micrographic surgery (MMS).1 Tumor recurrence rates, prolonged surgical times, and significant patient morbidity could be improved with tools that would allow for more accurate tumor evaluation and planning before surgery.
Several noninvasive imaging modalities have been applied to NMSC. Among them, high-frequency ultrasound (HFUS) can provide structural information about tumor size with its high resolution (∼50 μm) and deep penetration depth (>2 mm).1,2 However, the technique relies on mechanical contrast rather than functional contrast. It is desirable to increase tumor contrast for improved tumor demarcation. In this respect, optical contrast can complement ultrasound contrast. Photoacoustic imaging (PAI) can aid HFUS with its high contrast and sensitivity measures. Photoacoustic imaging is based on pulsed laser light being absorbed by chromophores (e.g., hemoglobin in blood), which leads to thermoelastic expansion and generation of sound waves.3 Photoacoustic imaging can quantify high-resolution vasculature with an achievable spatial resolution of ∼50 μm at ∼3 mm depth,3,4 thus being suitable for noninvasive imaging of skin malignancies.
This pilot study was designed to validate tumor thickness measurements obtained with HFUS by comparing them with histological thicknesses obtained during MMS. The authors also obtained vascular maps by PAI to enhance contrast for improved tumor demarcation and to help establish these techniques for future clinical trials involving surgery.
Experimental Design and Methods
An IRB-approved clinical trial (protocol #I226912) was initiated at Roswell Park Cancer Institute.5 Twenty-one patients with biopsy-proven NMSCs, of a minimum diameter of 0.5 to 1 cm, who were scheduled for MMS were enrolled, and written informed consent was obtained from all subjects.
A commercial grade HFUS imaging system (40 MHz, Episcan; Longport Inc.) was used because of its capability of providing high spatial resolution (∼45 μm axial [depth] resolution) with an ∼5 mm signal penetration depth. Custom clinical PAI system had a tunable laser (5 nanosecond pulse width, 20 Hz pulse repetition rate), a 20 MHz focused ring transducer with a 600 μm multimode fiber so that the laser fiber and transducer were set coaxially in a combined probe. The laser beam was collimated with a beam diameter of ∼2 mm on the sample. The system had ∼5 mm signal penetration depth and ∼80 μm depth resolution. High-frequency ultrasound and PAI scans were performed sequentially. First, ultrasound gel was applied to the tumor and the HFUS probe was positioned at the center of the tumor. The B-scans were viewed in real time to visualize the tumor, and images were saved to the computer. Then, the PAI probe was positioned at the same HFUS probe position and raw line scans were acquired. Photoacoustic imaging images were obtained by postprocessing with custom software using MATLAB (MathWorks Inc.).
The patients then underwent MMS. The excised tumor layer for each NMSC was brought to the laboratory, and frozen sections were prepared and stained with hematoxylin & eosin (H&E) for morphological assessment. Histological tumor thickness measurements were performed by the Mohs surgeon (N.C.Z.).
Squamous Cell Carcinoma
Figure 1 shows results from a patient with a squamous cell carcinoma tumor. The H&E staining image showed a tumor extent with a largest measured thickness of 0.625 mm. The HFUS image showed a hypoechoic area defining a tumor of 0.73 ± 0.03 mm thickness (mean ± standard deviation of 3 independent measurements) (Figure 1A), and the PAI map at 580 nm showed enhanced optical absorption contrast (Figure 1B). The hemoglobin absorption at 580 nm is close isosbestic point (∼584 nm), thus the PAI signal is mainly originated from blood absorption in the microvasculature and the signal intensity is proportional to the blood volume.
Basal Cell Carcinoma
Figure 2 shows results from a patient with a nodular basal cell carcinoma tumor. The largest tumor thickness measured by the H&E was 0.50 mm, and HFUS image (Figure 2A) indicated a tumor size of 0.53 ± 0.02 mm. Figure 2B shows the PAI map, indicating high hemoglobin absorption in the tumor areas supporting the HFUS image.
High-Frequency Ultrasound Thickness Versus Histological Thickness: Statistical Results
High-resolution HFUS measurements were performed for all patients (N = 21). The statistical analyses were performed using MATLAB software. This analysis showed that there was a strong correlation (r 2 = 0.87) between ultrasound thickness and histological thickness, and the paired t-test showed no significant difference (p = .31) between 2 measurement groups.
To better guide surgery and therapy of NMSC, precise information about tumor depth and thickness is desired. Any additional contrast that can complement the structural contrast would be beneficial for improved visualization before management. The authors showed that HFUS provided high-resolution structural images such as thickness of NMSCs, which correlated well with the histological measurements. Photoacoustic imaging provided high optical contrast originated from the hemoglobin absorption. This was a pilot study to establish these techniques and parameters for the future clinical trials. The limitations of the study include involvement of a small number of patients, especially for PAI measurements. Photoacoustic imaging is a relatively new technique, and pilot clinical studies may help increased acceptance in clinical settings. The results suggest that a concurrent multimodal imaging system based on HFUS and PAI can assist surgeons in preoperative assessment of tumor characteristics and provide improved tumor demarcation during surgical planning.
Dedicated to Janet Morgan. The authors thank Anne Paquette and Ken Keymel for the help at the clinic.
1. Marmur ES, Berkowitz EZ, Fuchs BS, Singer GK, et al.. Use of high-frequency, high-resolution ultrasound before Mohs surgery. Derm Surg 2010;36:841–7.
2. Mogensen M, Nurnberg BM, Forman JL, Thomsen JB, et al.. In vivo thickness measurement of basal cell carcinoma and actinic keratosis with optical coherence tomography and 20-MHz ultrasound. Br J Dermatol 2009;160:1026–33.
3. Maslov K, Stoica G, Wang LV. In vivo dark-field reflection-mode photoacoustic microscopy. Opt Lett 2005;30:625–7.
4. Wang PH, Luh JJ, Chen WS, Li ML. In vivo photoacoustic micro-imaging of microvascular changes for Achilles tendon injury on a mouse model. Biomed Opt Express 2011;2:1462–9.
5. Rohrbach D, Muffoletto D, Huihui J, Saager R, et al.. Preoperative mapping of nonmelanoma skin cancer using spatial frequency domain and ultrasound imaging. Acad Radiol 2014;21:263–70.